Phylogeography of Neotropical trees

The Neotropics harbour some of the most species-rich forests in the world, and understanding their biogeographical history has major implications for debates about the origins of tropical diversity (Hoorn et al., 2010; Rull, 2011), for the success of conservation efforts (Faith et al., 2010), and for predicting responses to climate change (Willis et al., 2010). A variety of factors, acting on different time-scales, has shaped these forests and has left imprints in the genetic diversity of the tree species that define them. Phylogeography, the analysis of geographical structure in genetic variation within and among closely related species, provides a framework for relating evolutionary divergence to drivers such as ecological, climatic and geological change. In particular, by combining data from different parts of the genome, the influence of processes acting on different time-scales can be distinguished. In tackling the evolutionary history of understudied tropical plant species, phylogeography faces many challenges – biological, technical and practical. In some cases species may have poor taxonomic resolution or deep intraspecific genetic divergence, which demands analytical approaches from systematics as well as from population genetics. Widely distributed species may suffer homoplasy at useful loci or – for plants – potentially useful markers (located in the chloroplast genome) may show little or no diversity. Practically, the difficulty of collecting widely across a species’ range presents a major obstacle and may require the use of difficult material such as preserved herbarium voucher specimens. Yet the potential for phylogeographical analysis to deliver new insight justifies the effort, and many of the hurdles are now being overcome by technological advances and by collaborative international research efforts. In this Special Issue we present a series of original papers that address these challenges and delve into the genomes of trees to try to answer some longstanding questions in Neotropical biogeography. Neotropical forests have experienced major change on palaeogeographical (e.g. Neogene and earlier) as well as more recent (e.g. Quaternary) time-scales, but the relative influence of these drivers continues to be actively contested (Hoorn et al., 2010; Rull, 2011). For example, although the formation of much Neotropical diversity was once thought to have been primarily influenced by Pleistocene climatic change, many lowland tree species show Neogene patterns of genetic divergence, suggesting that older forces have played a pervasive role (Dick et al., 2013). The continental movements that gave rise to the modern Andes and the formation of Central America have been a major driver in Neotropical biogeography. The rising Andes reversed the course of the proto-Amazon river, which had previously flowed westwards (Hoorn et al., 2010), and generated a permanent rain shadow in the western Amazon. Their eroding mineral base enriched the sandy Precambrian soils of the Guiana and Brazilian plates and increased regional edaphic heterogeneity (Fine et al., 2013). The northern Andes are a barrier for rain forest species because of their high elevation, over which most lowland organisms are incapable of dispersing or reproducing (Janzen, 1967), and because the eastern cordilleras intersect with llanos savanna, which is generally too dry to support rain forest vegetation. Despite the enormity of the Andean–llanos barrier, there are a remarkably large number of lowland plant species with a crossAndean disjunction. In the small country of Ecuador, for example, 30% of the known vascular plants (c. 1438 species) have been collected in lowlands both east and west of the Andes (Jørgensen & Le onY anez, 1999). It has been suggested that these species were widespread prior to the uplift of the Andes (vicariance hypothesis), which implies a Neogene age and morphological stasis of a large proportion of Neotropical plant species (Raven, 1999). An alternative view is that the Andes have behaved as a geographical filter, or gatekeeper, for intercontinental species migration. For example, Gentry (1982) noted that Central American forests often harboured the single ‘weedy’ species of Amazonian tree genera. The characteristics of weedy species, being fast growing, ecological generalists, and able to disperse long distances, would have allowed them to traverse geographical barriers and become established during the climatic fluctuations of the Pleistocene, perhaps when the llanos region was wetter or when coastal corridors were more extensive due to lowered sea levels (Poelchau & Hamrick, 2013a). In this issue, Cavers et al. (2013), Rymer et al. (2013) and ScottiSaintagne et al. (2013a,b) examine phylogeographical structure in lowland cross-Andean tree species. Of these, Cordia alliodora perfectly fits the Gentry model. This species, which the authors sampled in 25 countries, provides high quality timber yet grows quickly across a range of habitats, including relatively dry forests. Phylogenetic inference supported a South American origin, and although the species shows its greatest genetic divergence across the Andes, shared cross-Andean haplotypes were taken to indicate Quaternary dispersal. Similarly, the prominent ‘big-bang’ flowering tree Jacaranda copaia showed evidence of Amazonian origins in its pattern of intraspecific genetic and taxonomic variation, with relatively recent dispersal into Central America. In contrast, the Cedrela odorata species complex showed much deeper divergence and cryptic speciation, with a cross-Andean divergence pattern suggesting vicariance or Neogene dispersal. Because widespread Neotropical tree populations can be so old (Dick et al., 2013), it is possible that contemporary phylogeographical structure developed in landscapes that have since disappeared. Poelchau & Hamrick (2013a), for example,